Spark and Shark High- Speed In- Memory Analytics over Hadoop and Hive Data Matei Zaharia, in collaboration with Mosharaf Chowdhury, Tathagata Das, Ankur Dave, Cliff Engle, Michael Franklin, Haoyuan Li, Antonio Lupher, Justin Ma, Murphy McCauley, Scott Shenker, Ion Stoica, Reynold Xin UC Berkeley spark- project.org UC BERKELEY
What is Spark? Not a modified version of Hadoop Separate, fast, MapReduce- like engine» In- memory data storage for very fast iterative queries» General execution graphs and powerful optimizations» Up to 40x faster than Hadoop Compatible with Hadoop s storage APIs» Can read/write to any Hadoop- supported system, including HDFS, HBase, SequenceFiles, etc
What is Shark? Port of Apache Hive to run on Spark Compatible with existing Hive data, metastores, and queries (HiveQL, UDFs, etc) Similar speedups of up to 40x
Project History Spark project started in 2009, open sourced 2010 Shark started summer 2011, alpha April 2012 In use at Berkeley, Princeton, Klout, Foursquare, Conviva, Quantifind, Yahoo! Research & others 200+ member meetup, 500+ watchers on GitHub
This Talk Spark programming model User applications Shark overview Demo Next major addition: Streaming Spark
Why a New Programming Model? MapReduce greatly simplified big data analysis But as soon as it got popular, users wanted more:» More complex, multi- stage applications (e.g. iterative graph algorithms and machine learning)» More interactive ad- hoc queries Both multi- stage and interactive apps require faster data sharing across parallel jobs
Data Sharing in MapReduce HDFS read HDFS write HDFS read HDFS write iter. 1 iter. 2... Input HDFS read query 1 query 2 result 1 result 2 Input query 3... result 3 Slow due to replication, serialization, and disk IO
Data Sharing in Spark Input iter. 1 iter. 2... one- time processing query 1 query 2 Input Distributed memory query 3... 10-100 faster than network and disk
Spark Programming Model Key idea: resilient distributed datasets (RDDs)» Distributed collections of objects that can be cached in memory across cluster nodes» Manipulated through various parallel operators» Automatically rebuilt on failure Interface» Clean language- integrated API in Scala» Can be used interactively from Scala console
Example: Log Mining Load error messages from a log into memory, then interactively search for various patterns Base lines = spark.textfile( hdfs://... ) Transformed RDD RDD results errors = lines.filter(_.startswith( ERROR )) messages = errors.map(_.split( \t )(2)) tasks Driver cachedmsgs = messages.cache() Worker Block 1 Cache 1 cachedmsgs.filter(_.contains( foo )).count cachedmsgs.filter(_.contains( bar )).count... Result: scaled full- text to search 1 TB data of Wikipedia in 5-7 sec in <1 (vs sec 170 (vs sec 20 for sec on- disk for on- disk data) data) Action Cache 3 Worker Block 3 Worker Block 2 Cache 2
Fault Tolerance RDDs track the series of transformations used to build them (their lineage) to recompute lost data E.g: messages = textfile(...).filter(_.contains( error )).map(_.split( \t )(2)) HadoopRDD path = hdfs:// FilteredRDD func = _.contains(...) MappedRDD func = _.split( )
Example: Logistic Regression val data = spark.textfile(...).map(readpoint).cache() var w = Vector.random(D) Load data in memory once Initial parameter vector for (i <- 1 to ITERATIONS) { val gradient = data.map(p => (1 / (1 + exp(-p.y*(w dot p.x))) - 1) * p.y * p.x } ).reduce(_ + _) w -= gradient println("final w: " + w) Repeated MapReduce steps to do gradient descent
Logistic Regression Performance Running Time (s) 4500 4000 3500 3000 2500 2000 1500 1000 500 0 1 5 10 20 30 Number of Iterations 127 s / iteration Hadoop Spark first iteration 174 s further iterations 6 s
Supported Operators map filter groupby sort join leftouterjoin rightouterjoin reduce count reducebykey groupbykey first union cross sample cogroup take partitionby pipe save...
Other Engine Features General graphs of operators (e.g. map- reduce- reduce) Hash- based reduces (faster than Hadoop s sort) Controlled data partitioning to lower communication Iteration time (s) 200 150 100 50 0 PageRank Performance 171 72 23 Hadoop Basic Spark Spark + Controlled Partitioning
Spark Users
User Applications In- memory analytics & anomaly detection (Conviva) Interactive queries on data streams (Quantifind) Exploratory log analysis (Foursquare) Traffic estimation w/ GPS data (Mobile Millennium) Twitter spam classification (Monarch)...
Conviva GeoReport Hive Spark 20 0.5 Time (hours) 0 5 10 15 20 Group aggregations on many keys w/ same filter 40 gain over Hive from avoiding repeated reading, deserialization and filtering
Mobile Millennium Project Estimate city traffic from crowdsourced GPS data Iterative EM algorithm scaling to 160 nodes Credit: Tim Hunter, with support of the Mobile Millennium team; P.I. Alex Bayen; traffic.berkeley.edu
Shark: Hive on Spark
Motivation Hive is great, but Hadoop s execution engine makes even the smallest queries take minutes Scala is good for programmers, but many data users only know SQL Can we extend Hive to run on Spark?
Hive Architecture Client CLI JDBC Meta store SQL Parser Driver Query Optimizer Physical Plan Execution HDFS MapReduce
Shark Architecture Client CLI JDBC Meta store SQL Parser Driver Query Optimizer Cache Mgr. Physical Plan Execution Spark HDFS [Engle et al, SIGMOD 2012]
Efficient In- Memory Storage Simply caching Hive records as Java objects is inefficient due to high per- object overhead Instead, Shark employs column- oriented storage using arrays of primitive types Row Storage Column Storage 1 john 4.1 2 mike 3.5 1 john 2 3 mike sally 3 sally 6.4 4.1 3.5 6.4
Efficient In- Memory Storage Simply caching Hive records as Java objects is inefficient due to high per- object overhead Instead, Shark employs column- oriented storage using arrays of primitive types Row Storage Column Storage 1 john 4.1 1 2 3 Benefit: similarly compact size to serialized data, 2 mike but 3.5 >5x faster to john access mike sally 3 sally 6.4 4.1 3.5 6.4
Using Shark CREATE TABLE mydata_cached AS SELECT Run standard HiveQL on it, including UDFs» A few esoteric features are not yet supported Can also call from Scala to mix with Spark Early alpha release at shark.cs.berkeley.edu
Benchmark Query 1 SELECT * FROM grep WHERE field LIKE %XYZ% ; Shark (cached) 12s Shark 182s Hive 207s 0 50 100 150 200 250 Execution Time (secs)
Benchmark Query 2 SELECT sourceip, AVG(pageRank), SUM(adRevenue) AS earnings FROM rankings AS R, uservisits AS V ON R.pageURL = V.destURL WHERE V.visitDate BETWEEN 1999-01-01 AND 2000-01-01 GROUP BY V.sourceIP ORDER BY earnings DESC LIMIT 1; Shark (cached) 126s Shark 270s Hive 447s 0 100 200 300 400 500 Execution Time (secs)
Demo
What s Next? Recall that Spark s model was motivated by two emerging uses (interactive and multi- stage apps) Another emerging use case that needs fast data sharing is stream processing» Track and update state in memory as events arrive» Large- scale reporting, click analysis, spam filtering, etc
Streaming Spark Extends Spark to perform streaming computations Runs as a series of small (~1 s) batch jobs, keeping state in memory as fault- tolerant RDDs Intermix seamlessly with batch and ad- hoc queries map reducebywindow tweetstream.flatmap(_.tolower.split).map(word => (word, 1)).reduceByWindow( 5s, _ + _) T=1 T=2 [Zaharia et al, HotCloud 2012]
Streaming Spark Extends Spark to perform streaming computations Runs as a series of small (~1 s) batch jobs, keeping state in memory as fault- tolerant RDDs Intermix seamlessly with batch and ad- hoc queries tweetstream.flatmap(_.tolower.split).map(word => (word, 1)).reduceByWindow(5, _ + _) T=1 T=2 map reducebywindow Result: can process 42 million records/second (4 GB/s) on 100 nodes at sub- second latency [Zaharia et al, HotCloud 2012]
Streaming Spark Extends Spark to perform streaming computations Runs as a series of small (~1 s) batch jobs, keeping state in memory as fault- tolerant RDDs Intermix seamlessly with batch and ad- hoc queries tweetstream.flatmap(_.tolower.split).map(word => (word, 1)).reduceByWindow(5, _ + _) T=1 T=2 map Alpha coming this summer reducebywindow [Zaharia et al, HotCloud 2012]
Conclusion Spark and Shark speed up your interactive and complex analytics on Hadoop data Download and docs: www.spark- project.org» Easy to run locally, on EC2, or on Mesos and soon YARN User meetup: meetup.com/spark- users Training camp at Berkeley in August! matei@berkeley.edu / @matei_zaharia
Behavior with Not Enough RAM 100 Iteration time (s) 80 60 40 20 68.8 58.1 40.7 29.7 11.5 0 Cache disabled 25% 50% 75% Fully cached % of working set in memory
Software Stack Shark (Hive on Spark) Bagel (Pregel on Spark) Streaming Spark Spark Local mode EC2 Apache Mesos YARN